RIPS / IRSP, 19 (1), 87-131 © 2006, Presses Universitaires de Grenoble The Balance of Give and Take: Toward a Social Exchange Model of Burnout Abstract Based on the notion ofa disturbed balance between give and take, about a decade ago a research program on social exchange and burnout was initiated, Meanwhile, about twenty (longitudinal) studies have been carried out including more than eight thou- sand professionals such as teachers, physicians, nurses, police officers, prison officers, social workers, and mental disability workers. This article pulls together the results that have been obtained by the research program as well as by other researchers in the field. Based on empirical results, an integrative comprehensive social exchange model is proposed that includes three levels of social exchange with recipients, colleagues, and the organization as a whole. The hallmark of burnout - emotional exhaustion — appears to be related to lack of reciprocity at all three levels of social exchange. In addi- tion, professionals who feel disadvantaged at a particular level Donner et recevoir : une question d’équilibre. Vers un modéle du burnout fondé sur I’échange social Wilmar B. Schaufeli* Résumé A partir de Vidée d'un déséquilibre entre donner et recevoir, un programme de recherche surle lien entre échange social et burnout a été initié ily a une dizaine d’années. Une vingtaine d’études longitudi- nales ont été menées. Elles portent sur plus de huit mille professionnels tels que des enseignants, des méde- cins, des infirmiéres, des policiers, des gardiens de prison, des travailleurs sociaux et des personnes travaillant auprés de handicapés mentaux, Cet article rassemble les résultats obtenus au cours de ce programme de recherche et inclut ceux produits ailleurs par d'autres chercheurs. En se basant sur les résultats empiriques, un modéle intégrateur, compréhensif, de Péchange social est proposé, Il intégre 3 niveaux de l’échange social : avec les récipiendaires, avec les collégues et avec organisation dans son ensemble. Il apparait que ladimension centrale du burnout — épuisement professionnel —est lige au manque de réciprocité aux trois niveaux de l’échange social. De plus, * Department of Psychology and Research Institute Psychology & Health, Utrecht University, ‘The Netherlands. Correspondence should be addressed to Wilmar B, Schaufeli, PhD, Utrecht University, Department of Psychology, PO. Box 80140, 3508 TC Utrecht, The Netherlands, Phone: +31 30 2539216; Fax: +31 30 2537482; Email: WSchaufeli@fs.uu.nl REVUE INTERNATIONALE DE PSYCHOLOGIE SOCIALE 2006 N° 1 Key-words. Burnout, Equity, Reciprocity, Social Exchange Mots-clés Burnout, équité, réci- procité, échange socialof exchange tend to withdraw from that specific relationship. Furthermore, results indicate that simple and straightforward measures of reciprocity are to be preferred above more complicated measures, that is: (1) intraper- sonal measures of reciprocity that exclusively refer to one’s own stan- dards are superior to more complex interpersonal measures that include comparisons with others; (2) one-item ratings of reciprocity that are completed by the respondents are superior to researcher calculated ratio-scores les professionnels qui se sentent désavantagés a un niveau particu- lier de léchange tendent se reti- rer de cette relation spécifique. Par ailleurs, les résultats indiquent que les mesures simples et directes de la réciprocité doivent étre préférées aux mesures plus sophistiquées Plus précisément: 1) les mesures intra-personnelles de la réciprocité qui référent exclusivement a ses propres standards sont supérieures aux mesures interpersonnelles plus complexes qui incluent des compa- raisons avec autrui; 2) les auto-évaluations de la réciprocité of investments and outcomes basées sur un item sont supérieu- res aux scores calculés a partir du rapport entre les gains et les coats. ‘It is not the heavy emotional investment per se that drains the provider; rather it is an investment that has insufficient divi- dends’ (Heifetz & Bersani, 1983, p. 61) ‘It means that no matter how good you are, how much you put into this job, often you're just not going to reach the kids. I feel you put a lot more into your work than you get back. This real- ization is very depressing’ (A science teacher, cited in Blase, 1982, p. 105) From the outset, the a-theoretical nature of burnout research has been lamented. Burnout research has been criticized for being repetitive, non-innovative, data-driven, empiricistic, and a-theo- retical (Résing, 2003). But is burnout research really burned out, or is there still hope? By pulling together empirical results of a research program that started about a decade ago, the present article makes the case that a theoretical perspective from social exchange is fruitful in understanding the nature of burnout. BURNOUT AND SOCIAL EXCHANGEBurnout and social exchange As is illustrated by both quotations above, burnout has often been associated with some sort of imbalance or mismatch between investments and outcomes. Although the importance of lacking reciprocity in terms of a mismatch between investments and outcomes has been recognized throughout the burnout literature, this notion has not been explored systematically. At Jeast not until Buunk and Schaufeli (1993) made an attempt to link burnout with social exchange processes between caregivers and recipients. Their central thesis is that burnout develops primarily in the social and interpersonal context of the work orga- nization and that in order to understand its development attention has to be paid to the way individuals perceive, interpret, and construct the behaviors of others at work. Buunk and Schaufeli (1993) follow Maslach’s (1993) notion of burnout as a multidimensional syndrome that consists of emotional exhaus- tion, mental distancing (depersonalization and cynicism) and reduced personal accomplishment, which is rooted in the emotionally demanding interpersonal relationship between care- giver and recipient. By definition, this relationship is complementary in the human services, which is semantically well-illustrated by the terms ‘caregiver’ and ‘recipient’. Nevertheless, professionals look for some rewards in return for their efforts; for example, they expect the recipients of their care to show gratitude, to improve, or at least make a real effort to get well. However, these expectations often remain unfulfilled because, for instance, recipients do not improve as they suffer chronically, or they take the professional's efforts for granted. Hence, it is likely that over time a lack of reciprocity develops and that the balance between give and take is disrupted. As Buunk and Schaufeli (1999) have pointed out, reciprocity plays a central role in human life and establishing reciprocal social relationships is essential for the individual's health and well-being. They argue that the strong universal preference for reciprocal interpersonal relationships is deeply rooted since it may have fostered survival and reproductive success in our evolutionary past. In their theorizing, Buunk and Schaufeli (1993, 1999) draw heavily upon equity theory, probably the most influential social REVUE INTERNATIONALE DE PSYCHOLOGIE SOCIALE 2008 N° 1exchange theory. According to equity theory, people pursue reci- procity in social relationships: what they invest and gain from a relationship should be proportional to the investments and gains of the other party in the relationship (Adams, 1965). Moreover, when individuals perceive relationships as inequitable they feel distressed and they are strongly motivated to restore equity (Hatfield & Sprecher, 1984; Walster, Walster, & Berscheid, 1978). Or as Freudenberger and Richelson (1980, p.175) have put it: ‘Since burnout sets in when the effort spend is in inverse propor- tion to the reward received, it becomes imperative to balance the equation’. More specifically, Buunk and Schaufeli (1993) assumed that lack of reciprocity, or an unbalanced helping relationship, drains the professional’s emotional resources and eventually leads to emotional exhaustion ~ a hallmark of burnout. Initially, when the expected outcomes do not occur caregivers are likely to invest more effort in their relationships with recipients. When this does not pay off in terms of better outcomes, the imbalance increases and resources are further depleted, resulting a so-called ‘loss spiral’ (Hobfoll & Shirom, 2000). For most professionals, investing in a relationship without receiving appropriate ‘outcomes is a highly energy consuming, extremely depressing and frustrating. The resulting emotional exhaustion is typically dealt with by decreasing investments in the relationships with recipients. That is, by responding to recipients in a depersonal- ized way instead of expressing genuine empathic concern. Hence, according to Buunk and Schaufeli (1993) depersonaliza- tion — the second burnout component — can be regarded as a way of restoring reciprocity by withdrawing psychologically from recipients. Depersonalization involves a negative, callous, indif- ferent, or overly detached attitude to others. However, this way of coping is dysfunctional since it deteriorates the helping rela- tionship, increases failures and thus fosters a sense of diminished personal accomplishment — the third burnout component that is characterized by feelings of incompetence and doubts about one’s work achievements. Equity theory not only applies to the interpersonal level, but also to the organizational level where similar social exchange processes govern the relationship of employees with their orga- BURNOUT AND SOCIAL EXCHANGEnization. Therefore, Schaufeli, Van Dierendonck and Van Gorp (1996) have proposed a dual-level social exchange model that assumes that in addition to an unbalanced relationship at the interpersonal level, burnout is also caused by lack of reciprocity at the organizational level — that is, by a violation of the so-called psychological contract. The notion of psychological contract refers to the expectations held by employees about the nature of their exchange with the organization (Rousseau, 1995). It reflects the employees’ subjective notion of reciprocity: the gains or outcomes from the organization are expected to be proportional to one’s own investments or inputs. When the psychological contract is violated and reciprocity is corroded, this not only might lead to psychological distress (emotional exhaustion), but also a host of negative work outcomes is likely to occur, including the intention to quit, turnover, job dissatisfaction (Robinson & Rousseau, 1994), employee theft (Cropanzano & Greenberg, 1997), cynicism (Anderson, 1996), poor organizational commit- ment (Guzzo & Noonan, 1994), and absenteeism (De Boer, Bakker, Syroit & Schaufeli, 2002). From the perspective of equity theory, these responses to a violation of the psychological contract may be interpreted in terms of restoring reciprocity, either by increasing one’s outcomes (employee theft, illegitimate absenteeism) or by decreasing one’s investments: withdrawing psychologically (cynicism, reduced commitment) or behaviorally (turnover, absenteeism). In addition to the individual level and the organizational level, social exchange processes also play a role in work teams among colleagues. For instance, Buunk and Hoorens (1992) found some evidence that employees keep a ‘support bookkeeping’ that is based on the balance between giving and receiving support from others in their team. Given the centrality of the relationships with colleagues for work related outcomes, it seems plausible to expect that lack of reciprocity in the exchange relationship with one’s colleagues is an important determinant of burnout as well. Therefore, Taris, Schaufeli, Van Horn, and Schreurs (2004), added a third level at which social exchange processes may lead to burnout: the work team. They reasoned that team members who experience an imbalance between their investments in and their outcomes from the work team are likely to withdraw psychologi- REVUE INTERNATIONALE DE PSYCHOLOGIE SOCIALE 2006 N° 1Figure 1: Burnout at withdrawal at different levels of social exchange. cally by depersonalizing their colleagues in an attempt to restore reciprocity. Figure 1 summarizes the specific relationships between social exchange processes at three levels and their assumed outcomes that constitute the basis of our research program. Lack of reci- procity at all three levels of social exchange is expected to be associated with distress (emotional exhaustion), as well as with attempts to restore the balance of give and take at that specific level of exchange. This agrees with Lazarus and Folkman (1984) who distinguish between strains and coping behaviors. Strains - such as emotional exhaustion - may be considered generic outcomes, in the sense that they are likely to result from any disturbed exchange relationship. In contrast, coping behaviors — such as withdrawal — are tightly linked to particular exchange relationships. Lack of reciprocity at the interpersonal level Burnout Lack of reciprocity at the Withdrawal team level from colleagues Lack of reciprocity at the Withdrawal from organizational, = fF > the organization level Before discussing the findings of our research program on social exchange and burnout, it is imperative to make a brief note about the central concept of reciprocity. BURNOUT AND SOCIAL EXCHANGEReciprocity: concept and measurement In Adams’ (1965) seminal paper, the degree to which an exchange relationship is equitable is expressed in terms of the ratios of the investments and outcomes of one party and those of the other party, respectively. If one outweighs the other, inequity or lack of reciprocity exists, whereby equity is defined by as ‘..the equality of exchange between parties’ (Adams, 1965; p. 278). It has been correctly noted by Chadwick-Jones (1976, p. 234) that equity and reciprocity - as used in Adam’s original equity theory — are ‘almost identical terms’; both involve the comparison of the ratio of own investments and outcomes to that of another party. We prefer to use ‘reciprocity’ because this term is slightly more generic than ‘equity’, and because some of our opera- tionalizations differs from the classical equity formula (see below). Pritchard (1969) criticized the measurement of equity by means of Adam's (1965) classical formula because it neglects the role of ‘one’s internal standard as a means for comparison. This internal standard refers to ‘.. the amount of outcome Person perceives as being commensurate with bis own inputs, without regard to any comparison person’ (p. 205; Underscore in the original). According to Pritchard, rather than interpersonal comparisons as proposed in classical equity theory, intrapersonal comparisons play a crucial role in exchange processes. The internal standard is largely based on one’s past experience in similar exchange rela- tionships. Thus, in a sense intrapersonal comparison is a form of interpersonal comparison where the comparison other is substi- tuted by one’s own previous experiences. Following this lead, reciprocity is defined here as the equality of one’s perceived investments in and benefits from an exchange relationship, relative to the person’s own internal standards regarding this relationship. As several operationalizations of reciprocity exist an important issue is whether or not these measures can be used interchange- ably. Research on the concurrent validity of reciprocity measures is scarce and the results seem to be at odds with each other. For instance, whereas Prins, Buunk and Van Yperen (1993) asserted that different assessments of reciprocity led to similar results, REVUE INTERNATIONALE DE PSYCHOLOGIE SOCIALE 2008 N° 1Lujansky and Mikula (1983) reported poor intercorrelations between interpersonal and intrapersonal measures. Therefore, ‘one of the aims of the research program is to asses the concur- rent validity of various measures of reciprocity in predicting outcomes, most notably burnout (see the section on the measurement of reciprocity). Six different kinds of operational- izations of (lack of) reciprocity were used: 1. Adam's (1965) classical interpersonal equity formula, based on single ratings of one’s own and other's investments and outcomes. ‘Iypically, first some examples of investments in and outcomes from the relationship with, for instance, recipients are described (e.g. time, patience, effort, appreciation, and gratitude). Next, four questions are asked about the invest- ments in the relationship of both parties and about their outcomes; for instance; ‘Overall, how much do you feel you put into the relationship with recipients? and ‘Overall, how much do you feel recipients put into the relationship with you? (eg., Van Dierendonck, Schaufeli & Buunk, 2001). Instead of recipients, one’s colleagues may also act as compar- ison other: e.g, ‘How much do your immediate colleagues invest in the relationships with patients’ (e.g., Smets, Visser, Oort, Schaufeli & De Haes, 2004). 2. Adam's classical interpersonal equity formula, based on multiple-item scales for one’s own and other's investments and outcomes. For instance: ‘How much do you feel you invest in your work in terms of skills and energy? (own investments) and ‘How much do you feel you get in return from you work in terms of income and job benefits? (own outcomes) (¢.g., Taris, Kalimo & Schaufeli, 2002). 3. The intrapersonal ratio of one’s own outcomes relative to one’s own investments, based on a single rating of each. Sample items: ‘How much do you invest in the relationship with your patients? and ‘How much do you receive in return from this relationship? (e.g,, Smets et al., 2004). 4, Hatfield’s Global Measure of reciprocity (Hatfield, ‘Iraupman, Sprecher & Hay, 1985) that requests respondents to consider their investments in and outcomes from the relation involved, and then asks them to endorse the answer that best charac- terizes this relationship, using a seven-point rating scale. For BURNOUT AND SOCIAL EXCHANGEinstance, ‘The organization invests much more than it gains from me’ (+3); ‘The organization and I invest and gain equally’ (0); ‘invest much more in my work than Igain from the organization’ (-3). (¢g., Van Dierendonck, Schaufeli & Buunk, 1996). 5. Multi-item scales that assess intrapersonal lack of reciprocity and that include such items as: ‘J spend much time and consideration with my patients, but they gave me little appre- ciation back in return’ and ‘I benefit little from the effort I put in the organization’ (e.g., Schaufeli, et al., 1996). 6. The intrapersonal ratio of one’s own outcomes relative to one’s own investments, based on multiple-item scales. Sample items: ‘How much do you invest in motivating your students’ and ‘How much do you invest in coaching your students indi- vidually’ (investment in recipients); ‘How much appreciation do your students have for you’, ‘How much satisfaction do you get from your student’s personal growth’ (outcomes from recipients) (e.g., Van Horn, Schaufeli & Taris, 2001). A second issue of concern is whether inequity has curvilinear effects on the outcome variables rather than linear effects only. According to Adams (1965), receiving too little (underbenefited) as well as receiving too much (overbenefited) should result in negative outcomes. When the ratio between own investments and outcomes equals the ratio of others, the optimum level for the outcome variables is obtained (reciprocity). In intrapersonal approaches the optimum level is reached when one’s perceived investments equal one’s perceived outcomes. It might seem counterintuitive that human services professionals who feel over- benefited in their relationships with recipients may feel bad, but they often consider their job as a calling. Therefore, they are likely to be deeply personally involved in their relationships with recipients and, thus, as in intimate relationships, feeling over- benefited might induce negative feelings. Hence, in our research program we not only evaluated linear effects of lacking reci- procity, but also curvilinear effects (see the section on the measurement of reciprocity) REVUE INTERNATIONALE DE PSYCHOLOGIE SOCIALE 2006 N° 1‘Table 1 ‘Average proportion of perceived reciprocity (3). Is the balance between give and take disturbed in the human services? In order to answer this question we carried out a number of studies that are listed in Appendix 1. ‘Table 1 summarizes the results of Appendix 1 and shows the average proportions of (lack of) reciprocity across different samples and across different (intrapersonal and interpersonal) measures of reciprocity. Level ofsocial | k | i | N_ | Underbenefited | Balanced | Overbenefited exchange Recipients_| 5 | 8 [4077 36 37 7 Colleagues | 2 | 3 [3525 29 66 5 Organization | 4 | 7 [1,987 & 27 6 Note: Based on Appendix 1; k = number of samples; 1 = number of reciprocity indices ‘evaluated; N = total number of respondents, ‘Pwo general conclusions can be drawn. First, as ‘lable 1 shows, lack of reciprocity is most often experienced in the relationship with the organization, followed by recipients and colleagues, respectively. This trend is even more striking when different social exchange relationships within the same study are compared; for instance, Van Horn and Schaufeli (1996) found that 41% of the teachers in their sample felt underbenefited with regard to their students, whereas 76% felt underbenefited with regard to their school. As can be inferred from Appendix 1, comparable differences between interpersonal reciprocity and organizational reciprocity were found for medical specialists (Smets, et al., 2004), police officers (Kop, Euwema & Schaufeli, 1999)!, staff working with the mentally handicapped (Van Dierendonck, Schaufeli & Buunk, 1996), and nurses (Van Yperen, 1995) The only exception are therapists from a forensic psychi- atric center of whom 84% (sic!) felt underbenefited in their relationship with recipients, against ‘only’ 82% with the organiza- tion (Van Dierendonck, Schaufeli & Buunk, 1996). A minority of 10% of this sample felt that their investments and outcomes in the relationship with mentally disturbed criminal offenders were 1, Based on mean differences that are not represented in Appendix 1 BURNOUT AND SOCIAL EXCHANGEin balance - by far the lowest proportion. This might be explained by the fact that most criminal offenders suffer from severe personality disorders that are characterized by the very inability to reciprocate and thus maintain social relationships. Quite remarkably, the highest proportion of advantaged employees (24%) was found in the only non-human survives sample — a representative sample of the Finnish working popula- tion — which suggests that the balance of give and take is particularly disrupted among those who do people work. In conclusion: the balance between give and take is indeed disrupted among human services professionals, not only as far as the recipients and the colleagues are concerned, but particularly with regard to the organization. As will be discussed below, this result was likewise observed for intrapersonal and interpersonal measures of reciprocity. Lack of reciprocity and burnout ‘Table 2 summarizes the results of the studies on burnout and lack of reciprocity that are described in greater detail in Appendices 2- 4, Both the unweighted average correlation as well as the average correlation that is weighted for sample size are presented. Table 2: Level EEX DEP RPA ‘Average (un)weighted ofsocial | k | N correlations between exchange U yy U Ww U burnout and perceived Becipiencsyem 127) (11305 | 260 20m (mez el 7an Ole Iz eect Colleagues [ui | 8222] 20 [ 14 | 07 | 05 [| .09 | .07 | exchange. Organization | 17 | 9,599] 21 | 20 | 13 | 3 | a [12 Note: Based on Appendix 2; k = number of samples; N = total number of respondents; BEX = emotional exhaustion; DEP = depersonalization; RPA = reduced personal accomplish- ment; correlations are weighted for sample size; U = unweighted; W = weighted, Indeed, the mean correlations are strongest between emotional exhaustion and lack of reciprocity at ail three levels of social exchange, with the highest correlation for the interpersonal level of social exchange with recipients. Furthermore, and consistent with expectations, the correlation of depersonalization with reci- procity at the interpersonal level is stronger than with reciprocity at the team level and at the organizational level. Finally, the corre- lation between lack of personal accomplishment and reciprocity REVUE INTERNATIONALE DE PSYCHOLOGIE SOCIALE 2006 N° 1 Oyey at the interpersonal level is relatively high. Hence, lack of reci- procity at the interpersonal level is positively related with all three burnout dimensions: the less professionals feel that they get back from their investments in relationships with recipients, the more they feel exhausted, the stronger they depersonalize their recipients, and more their sense of accomplishment is diminished. It should be noted, though, that the average correlations in Table 2 are not very impressive. One possible explanation for these rather low correlations is that instead of linearly related, burnout and reciprocity are curvilinearly related (sce the next section). Furthermore, as can be seen from Appendices 2-4 the size of the correlations fluctuates considerably across samples from virtually zero to almost.70. In addition to sample effects, this large varia- tion might be due to different operationalizations of reciprocity. For instance, correlations with multi-item scales are consistently higher than with single-item measures or with ratio scores? (see the section on the measurement of reciprocity). Nevertheless, without any exception, all correlations are in the expected, posi- tive direction. In conclusion: anticipated, lack of reciprocity at all three levels of social exchange is positively related to emotional exhaustion, whereas, in addition, lack of reciprocity in the relationship with recipients is also positively related to depersonalization and reduced personal accomplishment. Reciprocity with recipients and burnout’ Of course, lack of reciprocity with recipients is not the only stressor that human services professionals face. In their review of the burnout literature Schaufeli and Enzmann (1998; pp. 81-85) discuss many other potential causes of burnout of whom work overload, role problems, and lack of social support seem to be 2, For reciprocity atthe interpersonal level average (unweighted) correlations of emotional ‘exhaustion, depersonaliztion and reduced personal accomplishment with multitem measu- res are 40, 38. and .28, respectively against 18, 13, and .14 with single-item measures and ratio-scores. For the organizational level the corresponding average correlations are: 36, 23, and 20, against 16, 11, and 10, respectively. 3. As the detailed account of research findings in the section below and in the section on reciprocity at various levels of exchange might be somewhat difficult to follow, the reader {s advised to consult Figure 4 that integrates and summarizes the main research findings, BURNOUT AND SOCIAL EXCHANGEthe most prominent. Hence, it is important to assess the strengths of the relationship between lack of reciprocity and burnout when such stressors are controlled for. In other words, does lack of reciprocity explain a significant and unique propor- tion of variance in burnout after the effects of work stressors or other relevant variables such as biographical characteristics are accounted for? Based on several studies, this question can be answered affirma- tive. For instance, in a sample of prison guards a significant effect of lack of reciprocity was found on ail three dimensions of burnout after controlling for workload, role conflict, and specific prison related stressors such as aggression and violence from inmates (Schaufeli, Van den Eijnden & Brouwers, 1994). In a similar vein, among teachers, reciprocity was significantly related to all three burnout dimensions, when controlled for age, gender, number of hours worked, and teaching experience (Van Horn, Schaufeli & Taris, 2001). Furthermore, in two samples of Polish and Dutch nurses, lack of reciprocity was associated with ail three levels of burnout, not only after controlling for work stressors (uncertainty and lack of control), but also for person ality (self-esteem and emotional reactivity), hours worked, and team size (Schaufeli & Janczur, 1994). The fact that similar results were obtained in both national samples lends further credence to the validity of the research findings. Moreover, it was shown in a sample of intimate partners, who cared for their spouses who suffered from either cancer or from multiple sclerosis, that the relationship between caregiver burnout and lack of reciprocity held when marital quality, gender, the duration of the illness, the physical and psychological condition of the ill partner, and support from other persons were controlled for (Ybema, Kuijer, Hagedoorn & Buunk, 2002). However, Van Horn and Schaufeli (1996) using a sample of primary and secondary teachers, failed to observe a significant relationship between lack of reciprocity with students and burnout. ‘Two years later their study was replicated with one important difference: instead of the researchers calculating the investments-outcome ratio, Hatfield's global, single-item measure was used which requires the teachers to asses their balance of give and take themselves (Peeters, Geurts & Van Horn, REVUE INTERNATIONALE DE PSYCHOLOGIE SOCIALE 2006 N° 11998). This time, lack of reciprocity was significantly and posi- tively related to all three burnout dimensions, also when controlled for age, gender, type of school, number of hours worked, and teaching experience. ‘Taken together, lack of reciprocity with recipients is positively related to burnout, also after controlling for various work stres- sors, interpersonal characteristics, personality characteristics, and demographic variables. Up to 10% of the variance in burnout is uniquely explained by lack of reciprocity; that is, after the effects of other relevant variables have been partialled out. Moreover, it seems that a global single-item rating of reciprocity yields better results than a similar ratio-score that is calculated by the researchers. All previous studies were all cross-sectional in nature. But what about longitudinal effects; does lack of reciprocity predict burnout over time? A longitudinal study among a representative sample of Dutch general practitioners spanning five years sheds light on this question. In the initial cross-sectional study at Time 1 (Van Dierendonck, Schaufeli & Sixma, 1994) a structural equa- tion model was tested that assumed that harassment by patients would lead to lack of reciprocity, which in turn would provoke emotional exhaustion followed by the development of negative attitudes (i.e. depersonalization and lack of personal accomplish- ment). Most importantly, it was reasoned that these negative attitudes would worsen the doctor-patient relationship and foster harassment by patients. In other words, a circular process was assumed: patient harassment — lack of reciprocity > emotional exhaustion + negative attitudes — patient harassment (see Figure 4). It appeared that this hypothesized model fitted quite well to the cross-sectional data. Five years later a follow-up was conducted so that the model could be studied longitudinally. As expected, it was found that negative attitudes towards patients at Time 1 increase the likelihood of being harassed by them five years later at Time 2, which fosters a lack of reciprocity and even- tually leads to burnout (Bakker, Schaufeli, Sixma, Bosveld, & Van Dierendonck, 1998). Thus, a lack of reciprocity in the caregiver recipient relationship seems to play an important role in the development of burnout through the impairment of the quality of the doctor-patient relationship. BURNOUT AND SOCIAL EXCHANGEAs previously noted, equity theory predicts a curvilinear or U- shaped relationship between reciprocity and burnout. Indeed, it was found among health care professionals that feeling deprived or underbenefited in the relationship with recipients as well as feeling advantaged or overbenefited results in higher future exhaustion levels, as measured at the one-year follow up (Van Dierendonck, Schaufeli and Buunk; 2001). In contrast, 770 indica- tion was found for a longitudinal (curvilinear) relation between reciprocity and depersonalization and reduced personal accom- plishment. The U-shaped relationship between reciprocity and emotional exhaustion that is depicted in Figure 2 has two rather unexpected characteristics, though. Emotional exhaustion a — —) eS 0 Underbenefited Balanced Overbenefited Note: Adapted from Van Dierendonck, Schaufeli & Bunk (1996, p.48). First, contrary to what equity theory predicts, rather than feeling underbenefited feeling overbenefited leads to higher future exhaustion scores. A similar relationship is also found in a cross- sectional study among therapists from a forensic psychiatric clinic (Van Dierendonck, et al., 2001). Secondly, and again contrary to what equity theory predicts, the lowest level of exhaustion was observed for the deprived or underbenefited relationship and not for the balanced relation- ship with patients. Obviously, an advantaged relationship runs REVUE INTERNATIONALE DE PSYCHOLOGIE SOCIALE 2006 N° 1 Figure 2; Curvilinear relation between reciprocity in the relationship with recipients (Time 1) and ‘emotional exhaustion ime 2),102 counter to the professional's attitude — which is directed at giving = in such a strong way that it might become stressful in itself. Recently, ‘Iruchot and Badré (2006) offered an interesting expla- nation for this intriguing result by discriminating between two different helping paradigms. The medical model that is predomi- nant in health care settings, assumes that patients are ‘victims’ who are supposed to accept their fate passively and follow the prescription of the expert. Hence, the investment of the patient in the relationship with the health professional is expected to be low. Patients are considered uncooperative when they resist their passive role. Health professionals might burn-out because the perceived investments of their active patients exceed their own investments. On the other hand, the compensatory model, that is predominant in social work settings, assumes that clients are ‘active agents’ who are responsible for the solution of their own problems. Hence, their investments in the helping relationship are expected to be high. Recipients are perceived as uncoopera- tive when they do not act as active agents. Social workers might burn-out when the investments of their passive clients are perceived to be low compared to their own investments. Using a vignet study among French nurses and social workers Truchot and Badré (2006) found that, as expected, nurses who were confronted with the overbenefice scenario experienced more burnout compared to nurses who were confronted with the underbenefice scenario. For social workers, the reverse was observed. As far as nurses are concerned, these findings agree with those of both studies by Van Dierendonck, Schaufeli, and Buunk (1996, 2001). Hence, it seems that in medical settings feeling overbenefited in the relationship with patients rather than feeling underbenefited may act as a risk-factor for developing burnout. In a study among hospital nurses, Van Yperen, Schaufeli and Buunk (1992) showed the moderating role of a particular person- ality characteristic. It appeared that nurses who perceived a lack of reciprocity in the relationship with their patients and who were low in communal orientation (i.e. do not have the desire to give and receive benefits in response to the needs of others) exhibited high levels on all three burnout dimensions. ‘This result was replicated in an independent study among hospice nurses BURNOUT AND SOCIAL EXCHANGE(Van Yperen, 1995), but not among medical specialists (Smets, et al., 2004). As illustrated by the results of Truchot and Deregard (2001) it is again important to take the type of helping paradigm into consideration. In their study, they observed the buffering role of communal orientation on burnout only among profes- sionals who endorsed the medical model but mot among those who adhered to the compensatory model. Under the medical model lack of reciprocity is not troublesome for those who are high in communal orientation - that is, who are responsive to the needs of others - because the investments of the patients are expected to be lower than the investments of the professional anyway. In that case low investments of patients go along with high investments of professionals. On the other hand, when communal orientation is low, lack of reciprocity is associated with high levels of burnout. One study investigated the relationship between lack of reci- procity and burnout within as well as outside the work setting (Bakker, Schaufeli, Demerouti, Janssen, Van der Hulst & Brouwer, 2000). It was found that among teachers lack of reciprocity in the intimate relationship with one’s partner was related to depres- sion (and not to burnout), whereas lack of reciprocity in the relationship with students was related to burnout (and only indi- rectly to depression). Accordingly, burnout and depression are both linked with similar social exchange processes, which, however, occur in different domains: burnout is related to lack of reciprocity in the professional domain and not in the private domain. In conclusion: lack of reciprocity at the interpersonal level is clearly and convincingly related to all three dimensions of burnout, even after controlled for a host of variables such as work stressors, interpersonal characteristics, personality characteris- tics, and demographics. Also, there is some evidence for the role of lack of reciprocity in the development of burnout, namely through a progressive deterioration of the quality of the care- giver-recipient relationship. Furthermore, reciprocity is curvilinearly related to burnout (i.e. exhaustion), both cross- sectionally as well as longitudinally. However, against expectations, instead of feeling underbenefited, feeling overben- efited is more strongly related to burnout, which seems to be REVUE INTERNATIONALE DE PSYCHOLOGIE SOCIALE 2008 N° 1specific for health care settings where the medical helping model prevails. A low level of communal orientation, or being less responsive to the needs of others, appears to be a protective factor for burnout, at least for nurses working in health care settings. Finally, the relationship between lack of reciprocity and burnout seems to be specific for the professionals working with other people in the human services. Reciprocity with the organization and burnout Several studies were conducted to specifically explore social exchange processes at the organizational level in relation to burnout. In a study among mental health care professionals, a structural equation model was successfully tested that assumed that lack of reciprocity with the organization affected both emotional exhaustion and the intention to leave the organization (Geurts, Schaufeli, & De Jonge, 1998). Both effects appeared about equally strong. Furthermore, lack of reciprocity with the organization seemed to follow from negative communication about management; the more negative the professionals rated their communication with management the more unbalanced their relationship with the organization. Finally, emotional exhaustion played a mediating role between lack of reciprocity and depersonalization (see Figure 4). Somewhat similar findings were obtained in a longitudinal study among teachers, again using structural equation modeling (Taris, Schaufeli, De Boer, Schreurs & Caljé, 2000): lack of reciprocity with the organization was associated with emotional exhaustion, psychosomatic complaints, poor organizational commitment, and future absen- teeism. Furthermore, it was observed among mental retardation staff and among therapists working with mentally disturbed crim- inal offenders (Van Dierendonck, Schaufeli & Buunk, 1996) as well as among nurses (Van Yperen, 1995, 1998) that lack of reci- procity with the organization is related to emotional exhaustion, but /ot to depersonalization. In short; these result suggest that lack of reciprocity at the organizational level seems to have two major consequences: emotional exhaustion and withdrawal from the organization. BURNOUT AND SOCIAL EXCHANGEVan Yperen (1998) demonstrated the role of self-efficacy as a moderator that ameliorates the negative effects of poor informa- tional support of maternity nurses on their perceived levels of reciprocity with the organization. That is, particularly nurses with low levels of self-efficacy experienced a lack of reciprocity with the organization when they received little information about organizational goals and policy issues, new equipment, new work processes, and service levels (see Figure 4). A study among a representative sample of the Finnish working population revealed that lack of reciprocity with the organization (partly) mediated the relationship between past and future downsizing on the one hand, and burnout and health complaints on the other hand (Kalimo, Taris & Schaufeli, 2003). Having expe- rienced downsizing in the past, or the anticipation of downsizing in the future, was associated with a disturbed balance between low work outcomes and high work investments. In its turn, this disturbed balance was associated with elevated levels of exhaus- tion, health complaints and reduced professional efficacy‘. In accordance with Geurts, et al. (1998) lack of reciprocity affected cynicism indirectly via exhaustion (see Figure 4). Another study using the same sample revealed that, instead of curvilinearly related, lack of reciprocity with the organization was linearly related with burnout (Taris, Kalimo & Schaufeli, 2002). More particularly, the underbenefited group showed the highest burnout scores, whereas the balanced group and the overbene- fited group exhibited similar but lower burnout scores. This result differs from the results among human services profes- sionals where the overbenefited group showed the most elevated burnout (emotional exhaustion) levels (see Figure 2). Hence, instead of a U-shaped relationship, as expected by equity theory and as observed among human services professionals (Van Dierendonck, et al., 1996; 2001), a J-shaped relationship is observed among the general working population: those who felt disadvantaged reported clevated levels of distress, whereas those 4, This study used the MBI-General Survey (Schaufeli, Leiter, Maslach & Jackson, 1996) that is equivalent to the original MBI, except that it also can be used outside the human servi- ces. The labels of the three MBI-GS scales have been slightly renamed in exhaustion (emotio- nal exhaustion), cynicism (depersonalization), and professional efficacy (personal accomplishment). REVUE INTERNATIONALE DE PSYCHOLOGIE SOCIALE 2006 N° 1 105who either felt in balance with their organization or who felt advantaged had lower distress scores, Last but not least, a series of studies have confirmed the organi- zational withdrawal hypothesis, showing direct as well as indirect effects of lack of reciprocity with the organization on various organizational outcome measures. For instance, a direct effect was observed on registered (future) absenteeism among mental health professionals (Geurts, Schaufeli & Rutte, 1999), bus drivers (Geurts, Schaufeli & Buunk, 1993), blue collar workers (Geurts, Buunk & Schaufeli, 1994; Van Yperen, Hagedoorn, & Geurts, 1996), and teachers (Taris, et al., 2000). Additional indi- rect effects were observed of a lack of reciprocity at the organizational level on absenteeism via health complaints (Taris, et al., 2000; De Boer, et al., 2002), conflicts with superiors (Geurts et al., 1993) and tolerant absence norms (Geurts et al., 1994). In addition to absenteeism, lack of reciprocity with the organization is also associated with resentment Geurts, et al., 1999), turnover intention (Van Yperen, 1995; Van Yperen et al., 1996; Geurts et al., 1999), and organizational commitment (Taris et al., 2000). These direct and indirect relationships are depicted at the bottom of Figure 4. In conclusion: in accordance with equity theory, lack of reci- procity at the organizational level is related to distress (emotional exhaustion and health complaints), as well as to behavioral with- drawal (sickness absenteeism) and psychological withdrawal (turnover intention, poor organizational commitment). In addi- tion, a disrupted balance of give and take with the organization seems to act as a mediator between work stressors (ie., poor informational support, and past or anticipated downsizing) on the one hand and distress and organizational withdrawal on the other hand. Finally, depersonalization seems either not affected by lack of reciprocity at the organizational level, or indirectly effected via emotional exhaustion. BURNOUT AND SOCIAL EXCHANGE